Deep Active Learning Model for Adaptive Pet Attenuation and Scatter Correction in Multi-Centric Studies Publisher



Shiri I¹ ; Sanaat A¹ ; Jafari E² ; Samimi R³ ; Khateri M⁴ ; Sheikhzadeh P⁵ ; Geramifar P⁶ ; Dadgar H⁷ ; Arabi H¹ ; Assadi M² ; Uribe C⁸ ; Rahmim A¹⁰ ; Zaidi H¹ Show Affiliations
Source: 2021 IEEE Nuclear Science Symposium and Medical Imaging Conference Record# NSS/MIC 2021 and 28th International Symposium on Room-Temperature Semiconductor Detectors# RTSD 2022 Published:2021

Abstract

Quantitative PET imaging requires multiple steps including attenuation and scatter correction (ASC), commonly carried out using anatomical images (CT, MRI). Clinical centers are equipped with different scanners and use different acquisition/reconstruction protocols, and these may vary over time in a given center. As new scanners and protocols emerge, deep learning model performance may deteriorate significantly, even for models built using large datasets, hence the need to update the models. The aim of the current study was to apply a deep active learning model for adaptive PET attenuation and scatter correction in multi-centric studies. We enrolled 1110 18F-FDG and 950 68Ga-PSMA PET/CT images from 3 and 4 different centers, respectively. We implemented a deep residual network architecture with 20 blocks for different-level feature extractions. First, the deep neural network was trained on 18F-FDG PET images. Since the radiotracer distribution is different from 68Ga-PSMA PET, we used body fine tuning transfer learning to transfer ASC knowledge between radiotracers. We build deep learning-based ASC models on 68Ga-PSMA and then applied active learning approaches to build center-specific ASC model. We trained a 2D deep neural network for direct generation of CT-based attenuation corrected PET images from non-attenuation-scatter corrected (NAC) images of 68Ga-PSMA patients. For model evaluation, voxel-wise mean error (ME), mean absolute error (MAE), relative error (RE%), absolute relative error (ARE%) and structural similarity index (SSIM) were calculated between ground truth CT-based attenuation/scatter corrected and predicted PET images using the deep learning algorithm.We achieved a ME of 0.22±0.05, MAE of 0.80±0.05, RE of 2.72±7.5%, ARE of 10.0±4.5% and SSIM of 0. 98±0.02 in the test set. Overall, we applied transfer learning to transfer knowledge between different PET radiotracers and built specific models for each center separately using active learning approaches. © 2021 IEEE.